III. HIGH SCHOOL GPA AND SAT SCORES AS PREDICTORS OF FRESHMAN ACADEMIC SUCCESS
In March of 1999, the CAS Admissions Committee analyzed admission standards for UH Hilo, comparing the grade performance of freshmen UH Hilo students at the end of their first semester with their high school GPA and SAT test scores. A copy of that report is appended to this document. This data compilation was repeated for data from fall, 2004 and fall, 2005 to examine whether there was any strong pattern of change since 1999.
The following figures show scatter plots of UHH grade point average vs. the high school performance statistics: high school GPA, SAT mathematics score, and SAT verbal score (SAT verbal scores were not available for fall, 2005). Plots from the 1998 data sets are reproduced for comparison. Examination of the each of the data sets across the 3 sample years shows that the overall patterns have not changed. The scatter plots were so similar to 1998 data that statistical analyses were not repeated, because conclusions and interpretations would not change from the appended 1999 report. Among students that were admitted, and for which UHH GPA data are possible, there are only very weak relationships to high school performance statistics. Thus there are many students that just meet the threshold of admission who do well at UHH. It is not apparent that high school GPA and SAT performance data could be used any differently to make a stronger discrimination among entering students of who would likely perform well at UHH.
W. Mautz
March, 1999
Analysis of UHH Admissions Statistics,
Class Entering Fall, 1998
Standards for admission to UH Hilo were evaluated using data for the UH Hilo freshman class entering Fall, 1998. UH Hilo grade point averages for the first semester in fall, 1998 (F98GPA), were compared to student high school grade point average (HSGPA) and SAT scores, both verbal (SATV) and mathematics (SATM). This particular population of students was admitted based on a standard called "sliding scale". Students were admitted if both HSGPA and the sum of sat scores (SATS = SATV + SATM) were sufficiently high or if a lower score in one measure were compensated by higher performance in the other measure. Details of this particular standard are described below. Performance data from the fall, 1998, class were used to evaluate the potential impact of increasing admission standards at UHH.
First semester performance at UHH (F98GPA) was positively and statistically significantly correlated with each of the three high school performance measures, HSGPA (r=0.50, p<0.001), SATV (r=0.29, p<0.001), and SATM (r=0.22, p<0.001). Scatter plots for these pairs of variables are shown in Figures 1-3. When run in a multiple regression analysis, an equation can be generated using high school performance variables to predict college GPA. This predicted grade point index (PGI) equation has the form: PGI= a(HSGPA) + b(SATV) + c(SATM) + d, where a, b, c, and d are constants. The multiple regression equation for these fall, 1998 data was PGI = 1.036(HSGPA) + 0.00126(SATV) + 0.00101(SATM) - 1.966. In this analysis, HSGPA was the most powerful predictor (p<0.001) followed, although not statistically significantly, by SATV (p<0.06), and SATM (p<0.16). The SAT scores lose apparent importance in the multiple regression analysis, because SAT scores are already highly correlated with HSGPA and with each other; thus they add little additional explanation of variation in UHH GPA. For example, Figure 4 shows a scatter plot of SATV vs SATM for which the correlation coefficient, r = 0.61, P<0.001. It is also possible that the first semester course load at UHH may not draw as much on mathematics skills as later semesters with the result that SATM has a diminished predictive power for early college GPA. These results are essentially the same as the results of Susan Brown's earlier analysis of student performance data from the class of fall, 1996.
Once established from an earlier base of data, the predicted grade point index could potentially be used to evaluate and determine admission status of future UHH applicants. To examine the impacts of this possibility, Figure 5 shows actual GPA performance of students plotted against predicted GPA index from the multiple regression on high school performance measures. Two points must be recognized at the onset of interpreting these data. 1) The data set is necessarily restricted to students who were admitted; there is no information from students who were rejected, some of whom could possibly have succeeded at UHH. 2) The data represent UHH GPA from only first semester; students doing poorly at first may improve in subsequent semesters. The first point is apparent on the Figure 5 scatter plot as there is a cutoff of data for predicted GPA Index values less than about 1.5. These students were generally not admitted based on the 1998 "sliding scale" admission standards. The striking feature of Figure 5 is the large variation in UHH GPA at the low end of predicted GPA index values. There are many students with predicted GPA's less than 2 who performed well at UHH, some earning better than a 3.0 GPA. This would argue against increasing admission standards, because a substantial set of applicants will be rejected that can perform well at UHH. Indeed, because of this large variation at the low end of the distribution, there were probably a significant number of students rejected from the applicant pool who would have achieved GPA's greater than 2.0. Any reduction of standards will also, of course, allow admission of more students who will not perform successfully at UHH. This is the tradeoff that must be balanced in setting standards. However, given the large variation in UHH GPA performance at the lower end of the distribution, it is clear that if an admission standard were based on a multiple regression predicted GPA index, the standard predicted GPA index should be set at less than 2.0 to avoid rejecting a large pool of successful students.
The sliding scale admission standard, which was actually applied in fall, 1998, uses the sum of SAT scores (SATV + SATM = SATS) and HSGPA. A 2 dimensional array of possible HSGPA and combined SAT verbal and mathematics score was constructed, and fields in the array of combinations of HSGPA and sum SAT scores were defined for which students were either admissible, eligible for consideration, or not admissible to UHH. The boundaries between these fields can be fit by linear equations so that the "sliding scale" method can be compared to other admission standards. These equations take the combined SAT verbal and mathematics scores (SATS) as input and set a minimum HSGPA for admission or for "eligible for consideration" status. This is essentially what one does by eye on the sliding scale array; an upper bound line fits one sawtooth boundary on the array and another line (equation) fits the other. The equations for the array field boundaries of fall, 1998 are:
Threshold for "eligible for consideration": HSGPA > -0.00123(SATS) + 3.521
Threshold for admissible: HSGPA > -0.00123(SATS) + 3.765
Of 275 students admitted in fall, 1998 with complete HSGPA and SAT verbal and mathematics data, 274 met the minimum sliding scale standard for "eligible for consideration" and of these, 256 met the standard for "admissible."
The fall 1998 freshman class can also be used to compare other possible admission standards. Questions are restricted to asking: from the pool of n=275 students admitted in fall, 1998, with complete data sets for HSGPA and SAT scores, how many fewer students would be admitted under another hypothetical standard. A number of years ago, a set of progressively increasing standards was proposed to be phased in by UH Hilo over the years 1996-2002. Shown below are these proposed new standards and the percentages of this fall, 1998, set of students who would be admitted under the listed standards:
Old standards in effect in 1992-1995: HSGPA=3.0 or (HSGPA=2.5 and combined SAT=800)
94% admitted
New Proposed Standards:
Phase I (1996-1998) HSGPA=3.0 or (HSGPA=2.5 and combined SAT=900)
90% admitted
Phase II (1998-2000) HSGPA=3.0 or (HSGPA=2.5 and combined SAT=900 with
min SATV=430 and min SATM=400)
88% admitted
Phase III (2000-2002) HSGPA=3.3 or (HSGPA=2.8 and min SATV=430 and
min SATM=400)
65% admitted
Phase IV (2002-2003) HSGPA=3.3 or (HSGPA=2.8 and min SATV=500 and
min SATM=500)
50% admitted
The present established standard is Phase II but with no minimum individual SAT score standards applied. Furthermore, in fall, 1998, the sliding scale standard was used. With our current pool of applicants, adoption of Phase II, III, and IV standards will have a successively more drastic impact on numbers of students admissible to UH Hilo.